What makes Snap Casts (Interesting)

Moderators: Torsten, Merlin, gordonjudd

What makes Snap Casts (Interesting)

Post Number:#11  Postby VGB » Mon Apr 16, 2018 8:47 am

Hi Dirk

This point is also interesting, it appears to show that changes to v-fly and v-loop do not follow each other immediately as described by the tension model. I think I see the delay also when I make the cast, do you think there are any other possible causes than a testing issue?

And whereas the rod tip’s acceleration reversal point coincides with the threshold of an acceleration plateau in markers 3 and 4, it precedes the threshold point of the loop’s acceleration plateau by one step...

This time step difference may be due to coarse steps used and requires calling for the magnification glass some time in future.


Regards

Vince
IANACI - There’s no such thing as absolutes
User avatar
VGB
 
Posts: 2638
Joined: Thu Jan 10, 2013 12:04 pm

What makes Snap Casts (Interesting)

Post Number:#12  Postby Dirk le Roux » Mon Apr 16, 2018 1:07 pm

VGB wrote:
So rod tip acceleration from here on is visibly impeded (by what?)


Dirk

I think that the rod tip acceleration is impeded by the combined inertia of the fly leg and rod tip. I think it’s another occurrence of the rod tip going backwards when we first accelerate the rod.


Hi Vince

I checked again and during the time after the instant shown, the rod is progressively less bent. It may be an occurrence of Graeme having begun the stop, so impeded by less input.

All the best, Dirk
Dirk le Roux
 
Posts: 329
Joined: Thu Sep 19, 2013 5:09 pm
Location: Pretoria, South Africa

What makes Snap Casts (Interesting)

Post Number:#13  Postby Dirk le Roux » Mon Apr 16, 2018 1:22 pm

Graeme H wrote:I'm interested to see slightly positive rod tip acceleration at the highlighted frame in the last image posted above. The rod is still bent, indicating there is still some downward force, even if it's not full power. The graph indicates the tip is about to stop accelerating in the opposite direction to the force I'm applying, so its downward velocity is constant at that exact point. I guess it's a function of the image frame chosen at that very point, because the next point on the plot has positive acceleration (rod tip slowing.)


Hi Graeme

There may be some downward force still applied by the rod tip at that point, but as you can see from the plot, the tip's downward acceleration at that instant has already ceased and is now downward deceleration/upward acceleration. The downward force displayed by the bend there is likely smaller than the upward force at the loop at that time.

Regards, Dirk
Dirk le Roux
 
Posts: 329
Joined: Thu Sep 19, 2013 5:09 pm
Location: Pretoria, South Africa

What makes Snap Casts (Interesting)

Post Number:#14  Postby Dirk le Roux » Sun Apr 22, 2018 5:32 pm

VGB wrote:This point is also interesting, it appears to show that changes to v-fly and v-loop do not follow each other immediately as described by the tension model. I think I see the delay also when I make the cast, do you think there are any other possible causes than a testing issue?

And whereas the rod tip’s acceleration reversal point coincides with the threshold of an acceleration plateau in markers 3 and 4, it precedes the threshold point of the loop’s acceleration plateau by one step...

This time step difference may be due to coarse steps used and requires calling for the magnification glass some time in future.


Hi Vince

I checked using smaller steps (down from 5 to 3 frame steps) and the results for that instant now correlate more closely:

Plateau Loop Front.JPG
Rod tip zero acceleration vs. loop front acceleration

Plateau M3.JPG
Rod tip zero acceleration vs. Marker 3 acceleration

Plateau M4.JPG
Rod tip zero acceleration vs. Marker 4 acceleration

I stress again that the footage quality is quite good but not perfect, and as one uses smaller steps down to 1 frame per step, the data gets progressively noisier. The programme is very sensitive and small differences in eye-balled position of markers, with the imperfect video capture, influence the noise. Let’s call this the step catch 22, which can be overcome by very high resolution, high shutter speed, very high fps footage.

With the footage available, I found using 3-frame steps gives detail without noise getting too wild, but still some roughness especially where markers round the loop. I eye-balled marker positions as accurately as possible and anybody is welcome to ask for and check my files.

Regards,
Dirk
Dirk le Roux
 
Posts: 329
Joined: Thu Sep 19, 2013 5:09 pm
Location: Pretoria, South Africa

What makes Snap Casts (Interesting)

Post Number:#15  Postby Dirk le Roux » Sun Apr 22, 2018 5:39 pm

Hi all

My previous posts showed roughly the very early instant at which the upward tension on the fly leg equals the downward force of the fly leg’s weight in a snap cast, as well as rod tip acceleration during the downward stroke and its interplay with acceleration at various point on the distributed line.

Now to leg velocities. For three different instances, I compare Y axis velocities of markers on the two legs at a region further away from the loop, as the step size with the available footage and angular changes near the loop, including where markers pass through the dolphin nose, produce noise (the step catch 22). A small degree of noise and thus roughness remains but let’s see.

While the loop front is on its way down:

a earth frame M2 M3 0.225s.JPG
Earth frame leg velocities, loop descending

The loop front’s travel velocity should be:

V_trav = (V_fl + V_rl)/2 = (3.32 + - 27.5)/2 = 12.09m/s

Let’s compare this answer with the loop front’s measured velocity:

a LF 0.225s.JPG
Loop front velocity, loop descending
a LF 0.225s.JPG (37.26 KiB) Viewed 1080 times

Tracker can by set to use a marker’s reference frame. Setting as the loop front’s frame we can check how the loop’s calculated propagation velocity compares to the measured leg velocities in the loop front’s frame:

V_prop = (V_fl – V_rl)/2 = (3.32 - - 27.5)/2 = 15.41m/s

Compared to the measured leg velocities in the loop front frame:

a loop frame M2 M3 0.225s.JPG
Loop frame leg velocities, loop descending

Not bad.
Continued …
Dirk le Roux
 
Posts: 329
Joined: Thu Sep 19, 2013 5:09 pm
Location: Pretoria, South Africa

What makes Snap Casts (Interesting)

Post Number:#16  Postby Dirk le Roux » Sun Apr 22, 2018 5:48 pm

Next the measured velocities while the loop is ‘frozen’:

b earth frame M5 M7 0.512s.JPG
Earth frame leg velocities, loop frozen

The loop’s travel velocity should be:

V_trav = (V_fl + V_rl)/2 = (10.57 + - 10.33)/2 = 0.12m/s

Compared to the loop front’s measured velocity:

b LF 0.512s.JPG
Loop front velocity, loop frozen
b LF 0.512s.JPG (38.38 KiB) Viewed 1080 times

And the calculation for the loop front’s propagation velocity:

V_prop = (V_fl – V_rl)/2 = (10.57 - - 10.33)/2 = 10.45m/s

Compared to the measured leg velocities in the loop front frame:

b loop frame M5 M7 0.512s.JPG
Loop frame leg velocities, loop frozen

Again, not bad.

Continued …
Dirk le Roux
 
Posts: 329
Joined: Thu Sep 19, 2013 5:09 pm
Location: Pretoria, South Africa

What makes Snap Casts (Interesting)

Post Number:#17  Postby Dirk le Roux » Sun Apr 22, 2018 5:54 pm

And measured velocities while the loop is climbing:

c earth frame M7 M8 0.675s.JPG
Earth frame leg velocities, climbing loop

The loop’s travel velocity should be:

V_trav = (V_fl + V_rl)/2 = (12.22 + - 6.07)/2 = 3.075m/s

Compared to the loop front’s measured velocity:

c LF 0.675s.JPG
Loop front travel velocity, climbing loop
c LF 0.675s.JPG (37.55 KiB) Viewed 1080 times

And the calculation for the loop front’s propagation velocity:

V_prop = (V_fl – V_rl)/2 = (12.22 - - 6.07)/2 = 9.145m/s

Compared to the measured leg velocities in the loop front frame:

c loop frame M7 M8 0.675s.JPG
Loop frame leg velocities, climbing loop

The light blue track M 8 from the last sequence is that of the last marker before the leader. This track and that of the loop front during this marker’s rise were rather more erratic, which I don’t ascribe to noise from the footage and suspect may be related to the grass surface and its interaction with the thin line taper and leader. It would be interesting to check a similar cast with a smooth floor surface.

With the next series I will attempt an overview of velocity changes between the two legs in relation to the loop’s descent and eventual climb.

Regards,
Dirk
Dirk le Roux
 
Posts: 329
Joined: Thu Sep 19, 2013 5:09 pm
Location: Pretoria, South Africa

What makes Snap Casts (Interesting)

Post Number:#18  Postby Dirk le Roux » Sun Apr 22, 2018 6:06 pm

Dirk le Roux wrote:The loop front’s travel velocity should be:

V_trav = (V_fl + V_rl)/2 = (3.32 + - 27.5)/2 = 12.09m/s


Hi all

Correction, the answer in the above calculation had a minus missing and it should read:

V_trav = (V_fl + V_rl)/2 = (3.32 + - 27.5)/2 = -12.09m/s

Regards,
Dirk
Dirk le Roux
 
Posts: 329
Joined: Thu Sep 19, 2013 5:09 pm
Location: Pretoria, South Africa

What makes Snap Casts (Interesting)

Post Number:#19  Postby VGB » Sun Apr 22, 2018 6:22 pm

Thank you Dirk, I have used industry standard equipment up to 5000 FPS, you always get tracker noise. I think your M8 track might be because the legs are not perfectly parallel. I look forward to the discussion on why the loop starts to accelerate .

Regards

Vince
IANACI - There’s no such thing as absolutes
User avatar
VGB
 
Posts: 2638
Joined: Thu Jan 10, 2013 12:04 pm

What makes Snap Casts (Interesting)

Post Number:#20  Postby Dirk le Roux » Sat Jul 28, 2018 7:07 pm

Hi everyone

At long last I had the time to complete consolidating and cross checking the data to obtain a continuous rod leg and fly leg velocity history (along Y axis, slanted parallel to the cast) from the snap cast footage Graeme kindly provided.

Tracker data from individual marker runs down each leg were pieced together in Excel. At higher leg velocities, combined with the rather far apart 1m markers spacing, some small velocity data gaps occurred between marker velocity tracks, which I bridged in Excel with its ‘fill series’ function, set to ‘trend’.

Velocity data taken at 2 footage frames per step (may show more subtle detail) and another set taken at 3 footage frames per step (may show less noise) were integrated.

Acceleration data from Tracker had bigger gaps between legs’ marker tracks, so was calculated from the combined velocity data and spot checked against measured longish Tracker marker runs.

Loop propagation velocity can be pieced together from Tracker marker tracks with the loop travel track set as reference frame, but rather was calculated from the already pieced together leg velocities. Accuracy of the calculated loop propagation velocity is supported by compared calculated loop travel velocity vs. measured loop travel velocity (see graph below), which match well enough. Calculated loop propagation velocity was also spot checked against measured longish runs.

“Downward” and “upward”, unless otherwise stated, are along the parallel-to-the-cast slanted Y axis.

Time scaling of the footage, and resulting measured duration, velocity and acceleration, is not entirely certain. Graeme shared the footage as 240fps frame rate, but at that rate the rod leg’s free fall acceleration (this time along a Y-axis perpendicular to ground) right towards the end of the sequence measures at average -16.2m/s², which is higher than acceleration by gravity. At 180fps the same free fall section measures at average -9.1m/s² acceleration. This appears to be closer to reality, with air drag likely accounting for the difference with gravity’s 9.8m/s². So, the data presented here is based on a best guess frame rate of 180fps. In absence of a filmed marble drop or such like more solid reference though, values given here should still be taken with a slight pinch of salt.

Here is the plot of measured fly leg, rod leg and loop front travel velocities throughout the cast, with the calculated (as previously mentioned) loop propagation velocity:

LEGS VELOCITY.jpg
VELOCITIES

Fly leg velocity data, and therefore loop travel and loop propagation velocities, ends at the point where the last line marker (just before the leader butt) enters the loop.

Note, the rod leg’s velocity and calculated loop propagation velocity reach their highest values at the instance that rod leg acceleration (see below) reaches zero.

And here is the loop front travel measured vs calculated velocities, as check for the validity of loop front travel calculation used:

LOOP TRAVEL MEASURED VS CALCULATED.jpg
LOOP TRAVEL VELOCITY - MEASURED VS. CALCULATED

continued...
Dirk le Roux
 
Posts: 329
Joined: Thu Sep 19, 2013 5:09 pm
Location: Pretoria, South Africa

PreviousNext

Return to Technical Analysis

Who is online

Users browsing this forum: No registered users and 1 guest